Aims: The purpose of this study was to examine the long-term effects of baclofen in a large cohort of alcohol-dependent patients compliant to baclofen treatment. Methods: A hundred patients with alcohol dependence, resistant to usual treatments, were treated with escalating doses of baclofen (no superior limit). Alcohol consumption (in grams) and craving for alcohol were assessed before treatment and at 3, 6, 12, and 24 months. Assessments were simply based on patients’ statements. The outcome measure was the consumption of alcohol, rated according to the World Health Organization criteria for risk of chronic harm. Results: While all patients were rated “at high risk” at baseline, approximately half of them were rated “at low risk” at 3, 6, 12, and 24 months. The sum of patients who were at “low risk” and at “moderate risk” (improved patients) was 84% at 3 months, 70% at 6 months, 63% at 1 year, and 62% at 2 years. The constancy of improvement over the 2-years was remarkable. The average maximal dose of baclofen taken was 147 mg/day. Ninety-two percentage of patients reported that they experienced the craving-suppressing effect of baclofen. Significant relationships were found between the amount in grams of alcohol taken before treatment and the maximal dose of baclofen required, and between the existence of a mental disorder and a lesser effect of baclofen. Conclusion: Baclofen produces an effortless decrease or suppression of alcohol craving when it is prescribed with no superior limit of dose. Potential limitations in the effectiveness of baclofen include the coexistence of a mental disorder, the concomitant use of other psychotropic drugs, a lack of real motivation in patients to stop drinking, and the impossibility to reach the optimal dose of baclofen because of unbearable side-effects (sometimes possibly related to too sharp a protocol of dose escalation).